Topic: What would it take to clean up the <gender>_focus tags?

Posted under Tag/Wiki Projects and Questions

Right now, the <gender>_focus tags (male_focus, female_focus, intersex_focus, ambiguous_focus) are massively undertagged - a search for solo_focus ~male_focus ~female_focus ~intersex_focus brings up only 87 pages of results, while solo_focus -male_focus -female_focus -intersex_focus brings up 750+.

What can be done about this?

There's a few obvious places where progress can be made - for instance, solo_focus -male_focus -female_focus -intersex_focus male/female turns up 512 pages of results, virtually all of which are clear instances of female_focus - clearly far too many for me to imagine trying to tag them by clicking, waiting for the page to load, scrolling down, copy-pasting in female_focus, hitting enter, etc., but I understand that privileged users have some more efficient way to add tags - would that make this tractable?

That leaves another 551 pages, according to solo_focus -male_focus -female_focus -intersex_focus -male/female.

solo_focus -male_focus -male/female -female_focus -intersex_focus female accounts for 263 of those, again, virtually all female. (faceless_females and disembodied_vaginas being much rarer than faceless_males and disembodied_penises, and disembodied_hands being male by default, so if a solo_focus image contains a female she's very likely the focus)

solo_focus -male_focus -female_focus -intersex_focus -male/female male/male turns up 116 more pages of results, virtually all of which are clearly male_focus.

And solo_focus -male_focus -female_focus -intersex_focus intersex turns up 30 more pages which are virtually all untagged intersex_focus.

Which leaves only 142ish pages that would require significantly more manual examination. (There's some overlap between those numbers, so it's a little more than 142, but not much - only a couple pages)

Updated by SnowWolf

Like most cleanup projects, it requires a bunch of people willing to dedicate quite a bit of time to making a furry art booru more searchable. Plus a few more experienced ones to deal with the tricky cases.

I personally cannot go through more than 10-20 posts before I get bored af and leave the rest :p

Updated by anonymous

MyNameIsOver20charac said:
Like most cleanup projects, it requires a bunch of people willing to dedicate quite a bit of time to making a furry art booru more searchable. Plus a few more experienced ones to deal with the tricky cases.

I personally cannot go through more than 10-20 posts before I get bored af and leave the rest :p

I think the key here is the interface. Say you had something like one of those slideshow add-ons that lets you search something and view it as a slideshow - imagine if you could choose up to 4 tags that might be applicable to the images in a search, assign 'em to A S D F, and just tap through the slideshow quickly deciding which tags belonged to each image. You might be able to tag about an image every couple of seconds - so, like, on the order of 1000 images per person-hour. Like, five people with really strong opinions about porn tagging putting in a couple hours each could tag ten thousand images. Something like this would work great for simple, extremely-quick-to-judge-in-most-cases stuff like "is this post male-focused or female-focused", "is this solo, solo_focus, duo, group, or zero_pictured", etc.

Updated by anonymous

Victoria_Oblong said:
I think the key here is the interface. Say you had something like one of those slideshow add-ons that lets you search something and view it as a slideshow - imagine if you could choose up to 4 tags that might be applicable to the images in a search, assign 'em to A S D F, and just tap through the slideshow quickly deciding which tags belonged to each image. You might be able to tag about an image every couple of seconds - so, like, on the order of 1000 images per person-hour. Like, five people with really strong opinions about porn tagging putting in a couple hours each could tag ten thousand images. Something like this would work great for simple, extremely-quick-to-judge-in-most-cases stuff like "is this post male-focused or female-focused", "is this solo, solo_focus, duo, group, or zero_pictured", etc.

I've had an idea for an application like this, but I am currently bogged down with other things going on. I would certainly work on it when I find the time, but do not expect anything.

The big problem that comes up when you are doing something like this is there is a 150 tag edit per hour limit. If it took two seconds to tag a post, you would blow through this limit in five minutes.

What a project like this would need is the real commitment of a privileged+ user, or someone willing to spend five minutes every hour for ten thousand hours (1,500,000 / 150).

Updated by anonymous

Idem said:
I've had an idea for an application like this, but I am currently bogged down with other things going on. I would certainly work on it when I find the time, but do not expect anything.

The big problem that comes up when you are doing something like this is there is a 150 tag edit per hour limit. If it took two seconds to tag a post, you would blow through this limit in five minutes.

What a project like this would need is the real commitment of a privileged+ user, or someone willing to spend five minutes every hour for ten thousand hours (1,500,000 / 150).

Oh, yikes, there's a 150 edit/hour limit? Fuck, that's brutal. I guess these are never getting fixed, then.

Updated by anonymous

SnowWolf

Former Staff

Victoria_Oblong said:
Oh, yikes, there's a 150 edit/hour limit? Fuck, that's brutal. I guess these are never getting fixed, then.

Members who exhibit that they would benefit from being able to excede that limit and are generally "good at tagging" gennnnerally get the ability to tag more and faster.

Victoria_Oblong said:
imagine if you could choose up to 4 tags that might be applicable to the images in a search, assign 'em to A S D F, and just tap through the slideshow quickly deciding which tags belonged to each image.

This is a neat idea, but it also makes me want to cry: I tend to get kinda dyslexic. I'd rather have an overlay to click.

Bonus points if I could have several things going on at once (for example, the *_focus, the assorted penetration tags, etc etc)

MyNameIsOver20charac said:
I personally cannot go through more than 10-20 posts before I get bored af and leave the rest :p

I like throwing on youtube on another monitor, personally.

Or I pick a search I actually like and don't mind browsing though. :)

Updated by anonymous

Genjar

Former Staff

Victoria_Oblong said:
Oh, yikes, there's a 150 edit/hour limit? Fuck, that's brutal. I guess these are never getting fixed, then.

150 edits/hour is the maximum speed that most users can reach without getting sloppy with their tagging. It's still more than two posts per minute, and just opening the posts, checking if the tag applies, and then adding it usually takes longer than that.

And privileged+ doesn't speed it up much. Because even though you get access to tag scripting, most posts still need to be checked. Unless you're working on something that can be reliably tagged from the thumbnail, such as black_and_white.

So yes, it's brutal. Which is why most tagging projects never get beyond the first 1%.

Updated by anonymous

SnowWolf

Former Staff

Personally, I tend to get caught up adding other tags too. I think most of us develop "pet tags" that we like to see added. For me, it's usually colors and markings. Or background stuff. (mountains, forest....)

That all said, though...

The amount of work involved is HUGE... but every little bit makes a different, y'know? Slap some of those tags into your blacklist

as in...

your search was solo_focus -male_focus -female_focus -intersex_focus male/female

add the following to your blacklist:

solo_focusmale_focusfemale_focusintersex_focus

then search something like horse male/female or horse oral male/female and just.... have at it. Every post makes a difference :) maybe you'll never finish it on your own, but you'll make a difference.

Updated by anonymous

  • 1